JOURNAL OF OPTOELECTRONICS AND ADVANCED MATERIALS Vol. 18, No. 5-6, May – June 2016, p.

589 - 594

Non-destructive characterization of dielectric materials

by microwave spectroscopy in LS band
M. T. JILANIa*, L. Y. CHEONGb , A. S. SAANDc , W. P. WENc , M. Z. U. REHMANd
a
College of Computing and Information Sciences, PAF Karachi Institute of Economics and Technology, Karach,i Pakistan
b
Department of Fundamental & Applied Sciences, University Technology PETRONAS, Tronoh, Perak Malaysia
c
Department of Electrical and Electronic Engineering, University Technology PETRONAS, Tronoh, Perak Malaysia
d
Department of Physics, COMSATS Institute of Information Technology, Park road, Islamabad, Pakistan

The paper describes the characterization of dielectric materials, by using a resonance based ring-resonator method. A
microstrip ring resonator (MRR) method is presented to measure effective dielectric constant (εreff) of composite laminate
materials. Multilayer structure of MRR is designed and simulated using Ansoft’s High Frequency Structure Simulation.
Comparative study of polytetrafluoroethylene (PTFE), ceramic and ceramic-PTFE composites have been carried out to
analyze the εreff along with its dependence on sheet thickness. From results, it is clearly observed that as the thickness of
overlay is increased there is significant change in effective permittivity, which is more noticeable for high-permittivity ceramic
material than low-permittivity PTFE material. Measured data is found in good agreement with the reported data.

(Received September 28, 2015; accepted June 09, 2016 )

Keywords: Dielectric constant, effective permittivity, microstrip ring-resonator, ceramics, laminates, PTFE.

1. Introduction Knowledge of these parameters is essential to

Electric properties of materials are now widely used in
various industrial applications. By utilizing these
properties, new advancements and trends are introduced in
agriculture, material characterization, chemical, geo-
sciences, bio-engineering and pharmaceuticals for
development of quality products. For the dielectric
properties the main parameter is the complex permittivity
(ε*) which describes the behavior when it is subjected to
an electromagnetic field which is defined by [1]
𝜀 ∗ = 𝜀′ − 𝑗 𝜀"
where, j = −1, ε′ is the real part called dielectric constant
or dielectric permittivity which represents the ability of
material to store energy when an external electric field is
applied and the ε″ is an imaginary part, called loss factor
which shows the energy dissipation of a material to an
external electric field [2]. The ratio between dielectric
constant (real part) and loss factor (imaginary part) of
permittivity describes the parameter called, tangent-loss or
dissipation factor. The tangent loss (tanδ) represent the
energy lost per cycle divided by energy stored per cycle
[3]
𝜀"
𝑡𝑎𝑛𝛿 =
𝜀′
The loss mechanism is based on another important
parameter that is conductivity (σ). Conductivity is the
ability of a material to conduct an electric current and it is
measured by conducting a known amount of current at
constant voltage [4].
understand the behavior of dielectric material in the
presence of high electromagnetic fields. This information
can be used to design high performance RF and
microwave circuit and components. The substrate
materials are the vital part of high frequency circuits. The
main driving force behind, the characterization of substrate
material is to improve high-speed digital signal
propagation. In addition to substrate materials, fabrication
of other microwave circuits often requires a low to
moderate permittivity multilayer materials. The multilayer
microwave laminate composite materials provide effective
(1) and low loss signal propagation throughout the circuit.
Ultimately, to design a good quality and high-speed
microwave circuit requires a rapid and high accuracy
method for the electrical characterization of printed circuit
substrates. In literature different methods are used for the
characterization of printed circuit substrate, ceramics,
laminate, thin-films and foam materials.
In order to determine the dielectric properties of any
material, no single technique can characterize all materials
over entire frequency band. A measurement with accuracy
for both types (lossy and low-loss) of materials is
challenging so for each band and their subsequent losses
require separate methods. Among various techniques using
microwaves; one of the most important technique is
resonance method. It is simple, cost-effective and fast (in
(2)
terms of time) even having higher Q-factor (about 250)
than conventional one. Application of ring resonators in
characterizing the dielectric properties of materials is not a
new idea, they have been used to investigate the moisture
content in soil [5] and meat [6], for characterization of
materials a higher-order mode ring resonator demonstrated
in [7]. Using ring-resonator dielectric measurements of FR-
4 substrate, plastic and clay are found very close (2.27%) to
 M. T. Jilani, L. Y. Cheong, A. S. Saand, W. P. Wen and M. Z. U. Rehman

published literature [8]. Measurements for printed circuit
board material using microstrip ring resonator shown very
accurate results over wide frequency band [9]. Sensing for
the small amount of liquid solvents presented in [10] and
found very useful to identify the impurities in the solution.
Results from this technique are validated with other
approaches and it shows that noticeable shift in the
resonant frequency and the changes in the quality factor are
the good indicators.
Therefore, in this study the aim is to utilize the
microstrip ring resonator technique for the evaluation of
thick-sheet composite materials. The effective dielectric
constant (εreff) is determined for different composite
materials and the obtained data is validated with the
reported literature.
By re-arranging the equation (5), the effective permittivity
of the ring-resonator can be obtained from
𝑛𝑐
𝜀𝑟 𝑒𝑓𝑓0 = ( )2 (6)
2 𝜋 𝑓0 𝑟𝑚
Once the effective permittivity of the unloaded
resonator is determined, the effective permittivity of the
loaded sample can be evaluated. For this purpose, same test
configuration of multilayer resonator is used that shown in
Figure 1. It can be seen that, substrate has a thickness is h1
with its dielectric constant εr1, the printed circuit of ring
and ground plane have thickness t, overlay testing sheet
material has dielectric constant of εr2 with the thickness h2
and on top of structure the air with dielectric constant εr3 of
1. Ultimately, determination of effective permittivity with
an overlay test material (that is the, loaded-resonator), can
be calculated by [11]

𝑓0 2
𝜀𝑟 𝑒𝑓𝑓1 = 𝜀𝑟 𝑒𝑓𝑓0 ( ) (7)
𝑓1

The evaluation of effective permittivity is widely

discussed in literature [13, 14]. A numerical model is
presented in [14], to evaluate the effective permittivity in
the presence of an overlay dielectric layer on microstrip
structure. The method is validated, both experimentally
and numerically, for a thick-sheet overlay. The expression
Fig. 1. Side view of proposed multilayer microstrip ring- for a line capacitance is given by
resonator
∞
1 1 [𝑓(β)]e2
= ∫ 𝑑(βℎ1 )
2. Theoretical analysis C πε0 Q2 𝜀𝑟1 𝑡𝑎𝑛ℎ(βℎ2 ) + 1
0 (𝜀𝑟1 + 𝜀𝑟2 𝑐𝑜𝑡ℎ(βℎ1 )) (βℎ1)
𝜀𝑟1 + tanh(βℎ2 )
Dielectric characterization using microstrip ring- (8)
resonator is based on a fact that, any change in effective
permittivity cause changes in resonator’s resonant where, εr1 is permittivity of substrate, εr2 is permittivity of
frequency [11]. In other words, the placement of overlay superstrate, h2 is thickness of a superstrate, and h1 is
dielectric layer modifies the air-boundary of a resonator substrate thickness, β is Fourier variable and f(β)/Q is the
that changes the effective permittivity, thereby shifts the Fourier transformation of charge distribution function f (x),
resonant frequency. Theoretically, the resonance is produce which is obtained from Yamashita’s variational method
when mean circumference of ring is equal to an integral of [13], it is given as
the guided wavelength [8, 12]
βw
2 π r = n λg for n = 1,2,3… (3) 𝑓(β) 𝑠𝑖𝑛 ( )
2
= 1.6 ( )
Q βw
where, n is the mode number, λg is the guided wavelength 2
and r is the mean radius of ring. The resonant frequency βw βw
2.4 βw 2 𝑠𝑖𝑛 ( ) 𝑠𝑖𝑛2 ( )
can be determined for n modes from this equation. Since, 2 4
+ 𝑐𝑜𝑠 ( )− + … (9)
λg is frequency dependent, thus for microstrip ring- βw 2 2 βw βw 2
( ) ( ) ( )
resonator it can be related to frequency by 2 ( 2 4 )

𝜆 1 𝑐 From equations (8) and (9), the capacitance C can be

𝜆𝑔 = = (4)
√𝜀 𝑒𝑓𝑓 √𝜀 𝑒𝑓𝑓 𝑓 evaluated, C0 can be determined by replacing permittivity
of both layers to permittivity of air. Thus, the effective
where, εeff is the effective dielectric constant and c is the permittivity can be evaluated by
speed of light. Considering (3) with λg will give
𝐶
𝑛𝑐 𝜀𝑒𝑓𝑓 = (10)
𝑓= for n = 1,2,3 … (5) 𝐶0
2𝜋 𝑟 √𝜀 𝑒𝑓𝑓
 Non-destructive characterization of dielectric materials by microwave spectroscopy in LS band

of ring to strengthen the electric field at ring edges. All

design parameters of microstrip ring resonator are listed in
Table-1. The fabricated prototype of two-port microstrip
ring resonator is shown in Fig. 2.

3.2 Measurement of substrate materials

To study the overlay dielectric materials and their

thickness effect on the permittivity, a microstrip ring
resonator with multi-layer configuration is used. The
structure of ring-resonator is designed in Ansoft’s High
Fig. 2. Fabricated prototype of two port microstrip ring- Frequency Structure Simulation (HFSS) software. After
resonator necessary setup procedures, the simulation is carried out up
to four periodic resonances. The simulation results are
Table 1. Microstrip Ring Resonator Design Parameters initially analyze in unloaded condition and subsequently,
with sample loaded resonator. For loaded resonator three
Parameter Design Value different dielectric materials used; glass microfiber
Roger RT/Duriod 5880 εr1 = 2.2 reinforced PTFE composites with a dielectric constant of
Substrate height h1 = 787 mm εrA = 2.1, hydrocarbon ceramic laminates with a dielectric
Dissipation factor (DF) tan δ = 9x10-4 constant of εrB = 3.55 and ceramic-PTFE composites with a
dielectric constant of εrC = 6.15 used as a test material, data
Copper thickness t = 17.5 μm
has taken form open literature [15]. A sheet material is
Frequency f = 1 GHz placed on the top of the ring-resonator. Different
Chrac. Impedance Zo = 50 Ω thicknesses of these samples are then simulated with the
Feed line length l = 35 mm range from 500μm to 1400μm with each step-increment of
PTFE Composites 100μm. The typical experimental setup is illustrated in
εrA = 2.1 Figure 3, in which two-port ring-resonator is connected
(Sample A)
Ceramic laminates with vector-network analyzer and accessible with personal-
εrB = 3.55 computer.
(Sample B)
Ceramic-PTFE composites
εrC = 6.15
(Sample C)

3. Experimental

3.1 Design and development of ring resonator

The lower band of microwave frequencies provides

some distinct features than millimeters waves. Therefore,
in this study for dielectric measurements the L-band
frequencies are used. The ring-resonator is designed on
fundamental frequency of 1GHz, which yields the radius
of r=34.85mm, considering equation (5). The Roger
RT/Duriod 5880 material substrate, with dielectric
Fig. 3. Illustration of microstrip ring resonator connected
constant of 2.2 and height of 787μm is used to design the
with vector-network analyzer
ring-resonator. The substrate permittivity, its height, its
dissipation factor, copper thickness and line impedance
from Agilent ADS LineCalc tool, width of annular ring
4. Results and discussion
and effective permittivity are calculated as w=2.40mm and
εeff = 1.875, respectively. A 35mm long feed line with the
In this section, effective permittivity of microwave
coupling gap of 200μm connected to the ring with an outer
laminate materials and the effect of their thicknesses are
radius of 34.85mm and the conductor thickness taken as
investigated. Simulated results of S11 and S12 parameters of
17.5μm. Width of feed line and ring strip calculated as
microstrip ring-resonator without overlay material are
2.40mm for the characteristic impedance of 50Ω using
shown in Figure 4. Periodic resonances with 1GHz
Agilent ADS software. The total size of the circuit is
interval are observed in unloaded condition. For the
141 × 87.5mm, as sufficient space retained to both sides
 M. T. Jilani, L. Y. Cheong, A. S. Saand, W. P. Wen and M. Z. U. Rehman

(a) (b)

Fig. 4. Resonance frequency response of microstrip ring-resonator without

the test material (a) S11 parameter (b) S12 parameter

(a) (b)

Fig. 5. Resonance frequency response of PTFE as overlay test material (a) S 11 parameter
(b) S12 parameter, as function of thickness

(a) (b)

Fig. 6. Resonance frequency response of ceramic as overlay test material (a) S 11 parameter
(b) S12 parameter, as function of thickness
 Non-destructive characterization of dielectric materials by microwave spectroscopy in LS band

(a) (b)

Fig. 7. Resonance frequency response of ceramic-PTFE as overlay (a) S11 parameter (b) S12
parameter, as function of thickness

unloaded resonator, it shows the periodic resonances occur

at 1GHz intervals. The 1st resonance occurs at 1.02GHz
while 2nd resonance is produced at 2.05GHz and then 3rd
resonance is at 3.06GHz and the last 4th resonance is at
4.07GHz. In obtained results, the shift in resonance
frequency is determined by considering the insertion loss
i.e. S12 parameter than return loss i.e. S11 parameter. This
is mainly due to the two port network, in which more
emphasize, is on appeared resonances, that are dependent
on dielectric overlay material.
In Fig. 5, results shown for glass microfiber reinforced
PTFE composites as dielectric overlay material with the
thickness vary from 500μm to 1400μm. For PTFE
composites with low permittivity and the low loss
characteristics (tanδ = 0.0009) it exhibit 1% frequency
shift in S12 parameter from its resonance frequency for
500μm thickness at 1st resonance whereas at 4th resonance
this frequency shift reached up to 2.5%. For the increased
thickness, with 1400μm sample at first resonance a 2%
frequency shift is observe while at fourth it becomes 1.3%.
For the entire range of thickness, there is small variation in
1st to last, 4th resonance. It is evident from the results
shown in Figure 6 for hydrocarbon ceramic laminates
material that, due to higher permittivity and lossy nature
(tanδ = 0.0021) of ceramics the frequency shift is higher
than PTFE. While using 500μm sample it exhibits 7% shift
at first resonance and this shift increases up to fourth
resonance with 8% shift. Similarly for thickness of
1400μm with 0.92GHz loaded resonant frequency it
shown around 8% drift from unloaded resonant frequency
at first resonance while at fourth resonant this shift goes up
to 8.8%. Ceramic materials show higher variation with
respect to thickness, after 2nd resonance variation is high
and at 4th resonance it’s become higher. It can be further
seen from Figure 7 that microwave laminates that are
developed from ceramic-PTFE composites has higher
frequency shift. When the combination of ceramic-PTFE
is used as a test material, for the 500μm sample thickness,
first resonance shows 9% shift from resonant frequency
and for the fourth resonance this shift is increased up to
10%.
Fig. 8. Measured effective permittivity of the PTFE,
ceramic and ceramic-PTFE composites materials.
When the sample thickness is increased up to 1400μm
the change in frequency shift has also been observed. At
the 1st resonance for the thick sample, frequency shift was
10.3% and similarly for the fourth resonance this shift
increased up to 14.5%. Ceramic-PTFE material has shown
higher variation than alone PTFE and ceramic materials.
In this case, as thickness increases, variation in resonance
is also significant in successive resonances. This can be
seen in S11 and S12 responses of all three samples, the
response (in dB scale) seems very lossy, the factors other
than test material are the, loose coupling between feed and
the ring, losses itself in substrate and the air-gap.
In the view of above results it is clearly observed that,
for the low-permittivity and low-loss dielectric overlay
materials, the obtained shift in frequency is minimal.
However, as high-permittivity and high-loss overlay
materials are used as overlay, the shift in resonance
frequency is further increased. Ultimately, measurements
are taken to validate the simulation results. The materials
that are commercially available from vendor (Roger
corporation), such as PTFE material that is RT5880,
ceramic material that is RO4000 and the ceramic-PTFE
 M. T. Jilani, L. Y. Cheong, A. S. Saand, W. P. Wen and M. Z. U. Rehman
material that is RT6006, are used for this purpose. The
measured effective permittivity is shown in Figure 8 as a
function of thickness. The high-permittivity and high-loss
material ceramic-PTFE (RT6006) exhibits higher variation
in effective dielectric constant εreff than its low-permittivity
and low-loss counterparts. However, variation in effective
permittivity is reduced for the thick sheets. The measured
data are found in good agreement with the data provided
by the Rogers Corporation [16].
6. Conclusion
This paper presents the dielectric characterization of
microwave composite laminate materials using the ring-
resonator method. During this study it is observed that a
material with low-permittivity and low-loss, such as
PTFE, exhibits small variation in resonant frequency as a
function of sample thickness, even this variation is not so
significant in its successive resonances. In contrary, the
moderate-permittivity and moderate-loss material such as
ceramic exhibited obvious changes in resonant frequency.
In addition to these materials, a high-permittivity and
high-loss material, such as ceramic-PTFE this variation
becomes significant. Even it shows much higher variation
in its successive resonances. Thus it can be conclude that
the effective dielectric constant of high-permittivity and
high-loss material is more sensitive to thickness than low-
loss materials and this change is more obvious on higher
frequencies.
References
[1] M. S. Venkatesh and G. S. V. Raghavan, The Journal
of The Canadian Society for Bioengineering (CSBE),
47, 7.15 (2005).
[2] Agilent-Technologies, "Basics Of Measuring The
Dielectric Properties Of Material," CA, USA2006.
[3] M. C. Giráldez, P. J. Fito, F. Toldrá, P. Fito, Physical
Sensors for Quality Control during Processing -
Handbook of Meat Processing:
A John Wiley & Sons, Inc., 2010.
[4] Shyam Narayan Jha, K. Narsaiah, A. L. Basediya,
Rajiv Sharma, Pranita Jaiswal, Ramesh Kumar, et al.,
Journal of Food Science and Technology,
48, 387 (2011).
[5] K. Sarabandi and E. S. Li, IEEE Transactions on
Geoscience and Remote Sensing, 35(5), 1223 (1997).
[6] Muhammad Taha Jilani, Wong Peng Wen,
Muhammad Zaka Ur Rehman, Muhammad Talha
Khan, L. Y. Cheong, International Journal of Applied
Electromagnetics and Mechanics, (accepted for
publication), 2015.
[7] R. Hopkins and C. Free, "Equivalent circuit for the
microstrip ring resonator suitable for broadband
materials characterisation," IET Microwave Antennas
Propagation, vol. 2, 2008.
[8] Nafaa M. Shebani and B. M. Khamoudi,
"Measurement of Dielectric Constant of Some
Materials Using Planar Technology," presented at the
IEEE Second International Conference on Computer
and Electrical Engineering, 2009
[9] J. M. Heinola, P. Silventoinen, K. Latti, M. Kettunen,
J. P. Strom, "Determination of dielectric constant and
dissipation factor of a printed circuit board material
using a microstrip ring resonator structure," in 15th
International Conference on Microwaves, Radar and
Wireless Communications, MIKON, 2004,
pp. 202-205 Vol.1.
[10] K. Saeed, A. C. Guyette, I. C. Hunter, R. D. Pollard,
"Microstrip Resonator Technique for Measuring
Dielectric Permittivity of Liquid Solvents and for
Solution Sensing," in Microwave Symposium, 2007.
IEEE/MTT-S International, 2007, pp. 1185-1188.
[11] P.A. Bernard and J. M. Gautray, "Measurement of
Dielectric Constant using a Microstrip Ring
Resonator," IEEE Transactions On Microwave
Theory and Techniques,, vol. 39, 1991.
[12] K. Chang, L. H. Hsieh, Microwave Ring Circuits and
Related Structures: John Wiley & Sons, 2004.
[13] E. Yamashita, IEEE Transactions on Microwave
Theory and Techniques, 16, 529 (1968).
[14] I. J. Bahl and S. S. Stuchly, IEEE Transactions
on Microwave Theory and Techniques,
28, 104 (1980).
[15] F. I. John, "Dielectric Constant and Loss Data,"
Technical Report AFML-74-250, Laboratory for
Insulation Research, Massachusetts Institute of
Technology (MIT), vol. III, 1977.
[16] Rogers-Corporation, "High Frequency Laminates and
Circuit Materials," (Products: RO4000,
RT6006/6010LM, RT5870/5880),2014.
____________________________
*
Corresponding author: mtaha.jilani@gmail.com